rhoai-model-catalog-sources

name	rhoai-model-catalog-sources
metadata	{"author":"rhoai3-demo","version":"1.0.0","platform-family":"rhoai","platform-baseline":"repo","ocp-baseline":"repo","skill-group":"RHOAI Platform"}
description	Use when documenting, reviewing, or administering Red Hat OpenShift AI model catalog source governance: viewing catalog sources in Model catalog settings, adding Hugging Face repository or YAML-file sources, enabling or disabling sources, previewing include and exclude model visibility patterns, managing or deleting sources, configuring the model-catalog-sources ConfigMap in rhoai-model-registries, default source boundaries, Hugging Face public/non-gated and connected-environment limitations, YAML catalog labels, and verification in AI hub -> Catalog. Do NOT use for model registry lifecycle, model deployment, dashboard feature visibility, or live cluster changes without the OpenShift safety guard.

RHOAI Model Catalog Sources

Use this skill to manage and govern OpenShift AI model catalog sources for the active product baseline in docs/PLATFORM_BASELINE.md.

Source Grounding

Read references/source-capture.md before using product configuration details. Official Red Hat documentation is product authority. This skill adapts the official model catalog source governance guide to this repo's GitOps review model.

Scope

This skill covers:

• dashboard governance from Settings -> Model resources and operations -> Model catalog settings
• source review for the default Red Hat AI, Red Hat AI validated, and Other catalog sources
• adding catalog sources from Hugging Face repositories or YAML files
• enabling and disabling catalog sources without deleting source configuration
• include and exclude model visibility patterns and Preview validation
• Hugging Face organization slug handling and public, non-gated source limits
• YAML catalog source metadata requirements and labels
• deleting non-default catalog sources
• configuring custom YAML catalog sources through the model-catalog-sources ConfigMap in the rhoai-model-registries namespace
• verification in AI hub -> Catalog

Use other skills for adjacent work:

• rhoai-model-catalog-workflows for data scientist and AI engineer catalog discovery, evaluation, registration, and deployment workflows
• rhoai-dashboard-customization for the disableModelCatalog dashboard flag and broader OdhDashboardConfig feature visibility
• rhoai-model-serving-platform for deploying catalog-selected models on KServe, vLLM, NIM, or related serving runtimes
• rhoai-model-management-monitoring for operating already deployed models
• rhoai-model-registry for model registry provisioning, database selection, permissions, and administrator lifecycle
• rhoai-model-registry-workflows for registering catalog-selected models into registries and operating model versions

Demo Policy

For this repo:

• Treat the model catalog as a governed discovery surface, not a trust stamp.
• Prefer Red Hat-provided catalog sources for default demo flows.
• Require explicit compliance and provenance review before exposing external Hugging Face or custom YAML catalog sources to users.
• Keep new sources disabled until source metadata, visibility patterns, and Preview output are reviewed.
• Use include patterns first to narrow the source and exclude patterns second to remove blocked models.
• Do not use Hugging Face catalog sources for disconnected demo environments.
• Do not model gated or private Hugging Face repositories through this feature.
• Manage long-lived custom YAML catalog sources through GitOps once active GitOps implementation exists.
• Keep source-controlled YAML catalog examples minimal and traceable to official docs or approved demo policy.
• Do not delete default catalog sources.
• Do not claim custom catalog source support on s390x.

Workflow

1. Confirm the active baseline in docs/PLATFORM_BASELINE.md.
2. Read references/source-capture.md and references/official-doc-extraction.md.
3. Decide whether the task is dashboard-based source governance or ConfigMap/GitOps-based custom source configuration.
4. For Hugging Face sources, confirm the organization slug, connected-cluster requirement, and public non-gated model boundary.
5. For YAML sources, confirm model definitions include model name, description, and deployment information.
6. Apply include and exclude visibility patterns, then use Preview before enabling a source.
7. Verify source status in Model catalog settings and model visibility in AI hub -> Catalog.
8. Validate with references/validation-checklist.md.

References

• references/source-capture.md
• references/official-doc-extraction.md
• references/validation-checklist.md
• examples/model-catalog-source-patterns.md